Overview
The standard retrospective — "what went well, what did not go well, what should we change" — is fine for the first few sprints. After that, it becomes a ritual that teams endure rather than value. People write the same sticky notes every two weeks. The real issues stay unspoken because the format does not create space for them.
This guide gives you five alternative formats, each designed to surface a different kind of insight. They are ordered from lightest (good for teams new to retros) to deepest (good for teams that need to address systemic issues). Each format includes exact timing, facilitator scripts, and tips for handling the uncomfortable moments that signal you are actually getting somewhere. If you want a printable agenda with action-item tracking, the Sprint Retrospective Template pairs well with any of these formats.
Who this is for: Scrum masters, product managers, engineering managers, or anyone facilitating team retrospectives.
Time required: 60-75 minutes per format (adjust timing notes as needed for your sprint cadence)
When to use which format:
| Format | Best for | Avoid when |
|---|---|---|
| Start-Stop-Continue | Teams new to retros; generating quick, actionable changes | The team has deep systemic issues |
| 4Ls (Liked/Learned/Lacked/Longed For) | Mid-maturity teams; balancing positive and constructive feedback | The team is in crisis mode |
| Sailboat | Teams that respond well to visual metaphors; identifying forces | The team is highly analytical and dislikes abstraction |
| Mad-Sad-Glad | Teams avoiding emotional topics; surfacing morale issues | Trust is very low (use anonymous surveys first) |
| Timeline | After a major release, incident, or long sprint; understanding cause and effect | Short sprints with low event density |
Format 1: Start-Stop-Continue (60 minutes)
Why this format works
Start-Stop-Continue is the most action-oriented retro format. Every sticky note maps directly to a behavior change. There is no "what went well" category that people fill with generic positives — every column drives toward doing something differently.
Setup
Draw three columns on the board: Start (things we should begin doing), Stop (things we should quit doing), Continue (things that are working and we should keep doing).
Facilitation script and timing
Silent writing (8 minutes):
"Write sticky notes for each column. One idea per note. Be specific: 'start doing code reviews within 24 hours' is actionable. 'Start communicating better' is not. Aim for at least one sticky in each column."
Place and read (5 minutes):
Everyone places their notes. Facilitator reads each one aloud. Cluster duplicates.
Dot voting (3 minutes):
Each person gets 3 dots. Place them on the items you think would have the biggest impact if we actually followed through.
Discussion — top 3 items (20 minutes, ~7 minutes each):
For each top-voted item:
"Who wrote this one? Can you give us a specific example from this sprint?"
Then to the room:
"Does anyone see a reason this would be hard to implement? What would need to change for this to actually happen?"
Commit to actions (10 minutes):
For each of the top 3 items, document:
Review last sprint's actions (5 minutes):
"Before we close, let's look at what we committed to last sprint. [Read each item.] Did we do it? If not, why not?"
This review is non-negotiable. Skip it, and retro actions become suggestions that everyone ignores.
Facilitator tips for Start-Stop-Continue
Format 2: 4Ls — Liked, Learned, Lacked, Longed For (60 minutes)
Why this format works
The 4Ls format creates psychological permission to express needs ("Longed For") and gaps ("Lacked") without framing them as complaints. "We lacked clear requirements" feels different from "requirements were bad" — even though both identify the same problem. This subtle reframe makes teams more willing to be honest.
Setup
Draw four quadrants on the board: Liked (what went well), Learned (what we discovered), Lacked (what was missing), Longed For (what we wish we had).
Facilitation script and timing
Silent writing (10 minutes):
"Fill each quadrant. Liked: what energized you this sprint? Learned: what did you figure out that you did not know before — about the product, the customer, or the process? Lacked: what resource, information, or support was missing? Longed For: if you could change one thing about how we work, what would it be?"
Gallery walk (5 minutes):
Everyone places their notes. Instead of reading aloud (which takes too long with four quadrants), have the team spend 3 minutes reading all notes silently, then 2 minutes asking clarifying questions on any note that is unclear.
Theme identification (10 minutes):
"What patterns do you see? I am looking for themes that span multiple quadrants. For example, if 'Liked: design review was great' and 'Lacked: design review for the other feature' both show up, that is a signal."
Circle or highlight the themes the room identifies. Aim for 2-3 themes.
Deep dive on top theme (15 minutes):
Pick the theme with the most energy in the room (you will feel it — people lean in, multiple people want to talk).
"Let's spend 15 minutes on [theme]. What specifically happened? What was the root cause? What would we need to change to fix it — and is that change within our control?"
Action items (10 minutes):
Same structure as Start-Stop-Continue: specific change, owner, success measure, check-in.
Close — share one "Learned" (5 minutes):
"Before we go, I want each person to share one thing from the Learned quadrant — yours or someone else's — that you want the team to remember."
This ending on a positive note (learning, not complaint) makes people more willing to come back for the next retro.
Facilitator tips for 4Ls
Format 3: Sailboat (75 minutes)
Why this format works
The Sailboat metaphor — wind (what pushes us forward), anchors (what holds us back), rocks (risks ahead), island (our goal) — makes abstract concepts tangible. Teams that struggle with traditional retro formats often find it easier to think in metaphors. It also naturally surfaces risks, which most retro formats ignore entirely.
Setup
Draw a simple sailboat on the board. Label four areas:
Facilitation script and timing
Define the island (5 minutes):
"Before we start, let's align on where the boat is heading. What is our goal for the next sprint — or the next quarter, if we want to think bigger? I want one sentence that everyone agrees on."
Write it on the island. If the team cannot agree on the goal, that is the retro finding right there.
Silent mapping (10 minutes):
"Fill in each area. Wind: what is helping us move toward the island right now? Anchors: what is slowing us down? Rocks: what risks do you see ahead that we have not talked about? Be specific."
Place and cluster (5 minutes):
Everyone places their notes. Cluster duplicates. Count the total stickies in each area.
Proportional discussion:
Here is where the Sailboat format differs from others: allocate discussion time proportionally to the number of stickies, with a minimum of 10 minutes per area.
Action items (10 minutes):
For anchors: pick one to address with a specific action, owner, and timeline.
For rocks: pick one to mitigate. Define the early warning sign that would tell you the rock is getting closer.
Facilitator tips for Sailboat
Format 4: Mad-Sad-Glad (60 minutes)
Why this format works
Most retro formats ask about process and output. Mad-Sad-Glad asks about emotions. This surfaces the human factors that drive team performance: frustration with a decision, disappointment about a missed opportunity, joy about a collaboration that worked. Empowered teams need emotional safety as much as they need good processes — and this format explicitly creates space for that.
Setup
Draw three columns: Mad (frustrated/angry), Sad (disappointed/deflated), Glad (happy/energized).
Facilitation script and timing
Frame the exercise (3 minutes):
"This retro is about how we felt this sprint, not just what happened. I want honest emotional reactions. There are no wrong feelings. If something frustrated you, put it in Mad even if you think it is 'small.' Small frustrations that go unspoken compound into big problems."
Silent writing (8 minutes):
"Write one sticky per emotion. Mad: what frustrated or angered you? Sad: what disappointed you or felt like a missed opportunity? Glad: what made you feel good about being on this team? Include at least one Glad — even in a tough sprint, something went right."
Simultaneous placement (2 minutes):
Everyone places at once. No reading aloud yet — just a visual scan.
Facilitator reads and groups (5 minutes):
Read each note aloud in a neutral tone. Do not editorialize. Group similar themes. Count the stickies in each column.
Process each column (30 minutes total, 10 per column):
Start with Glad — this sets a positive baseline and builds trust before the harder conversations.
"What patterns do you see in Glad? What specifically made these things work?"
Then Sad:
"Sad is often about missed potential — things that could have been great but were not. What got in the way? Was it within our control?"
Then Mad — the hardest column:
"Mad items are friction points. Some might be about people, processes, or decisions. Let's talk about the situation, not the individuals. For each theme: what happened, why did it frustrate you, and what would need to change?"
Critical facilitation moment: If someone's "Mad" item is about another person in the room, redirect from blame to behavior: "It sounds like the handoff process between design and engineering created friction. What would a better handoff look like?" This keeps the conversation productive without silencing the legitimate frustration.
Action items (10 minutes):
Pick one item from Sad and one from Mad. Define specific actions to prevent recurrence.
For Glad items: document what worked so the team can intentionally replicate it.
Facilitator tips for Mad-Sad-Glad
Format 5: Timeline Retrospective (75 minutes)
Why this format works
The Timeline retro is the best format for complex sprints, major releases, or post-incident reviews. Instead of asking "what went well," it asks "what happened, in what order, and how did we react?" This chronological approach surfaces cause-and-effect relationships that other formats miss. The feature that felt frustrating might trace back to a vague requirements meeting three weeks ago — and you only see that connection when you lay events out in sequence.
Setup
Draw a horizontal timeline across the full width of the board. Mark the start date (sprint start or project kickoff) and end date (today). Add any known milestones between them (sprint planning, mid-sprint demo, release date).
Above the timeline: Positive Events (things that went well)
Below the timeline: Negative Events (things that went poorly)
Facilitation script and timing
Individual timeline construction (12 minutes):
"Think about this sprint chronologically. From day one to today, what happened? Write one event per sticky — include the approximate date or day. Place positive events above the line and negative events below. Include decisions, discoveries, blockers, wins, and anything that changed the direction of our work."
Build the shared timeline (10 minutes):
Everyone places their events on the timeline. Facilitator reads left to right, clustering events that happened on the same day.
"Let's walk through this. [Date]: we had sprint planning — several people noted [theme]. Then on [date], [event]. What happened between these two events?"
Identify inflection points (10 minutes):
Look for moments where the timeline shifts from positive to negative (or vice versa).
"I see a cluster of negative events around [date]. What was different about that week? And here, around [date], things seem to improve. What changed?"
Mark inflection points with a star. These are the moments that determined the sprint's trajectory.
Root cause discussion (20 minutes):
For each major inflection point (limit to 2-3):
"Let's trace this back. [Negative event] happened on [date]. What led to it? What decisions or events earlier in the timeline made this more likely?"
Use the "5 Whys" technique — but stop at the actionable root, not the philosophical one. "Why did the release fail?" → "Because we did not test the integration" → "Because the integration environment was down" → "Because nobody owns the integration environment." That last one is actionable. Going further ("why does nobody own it?" → "because our org structure..." ) leaves the team's control.
Action items (10 minutes):
For each root cause identified:
Close — key learning (5 minutes):
"Looking at this entire timeline, what is the one thing you would tell past-you at the start of this sprint?"
Facilitator tips for Timeline
General Facilitator Tips (All Formats)
On psychological safety:
On follow-through:
On timing and cadence:
On remote retros:
On handling difficult moments: